1. Field of the Invention
The present invention relates to communication networks and, more particularly, to a method and apparatus for encapsulating services for transportation over metallic physical mediums.
2. Description of the Related Art
Data communication networks may include various routers, switches, nodes, hubs, proxies, and other network devices coupled to and configured to pass data to one another. These devices will be referred to herein as “network elements.” Data is communicated through the data communication network by passing data packets, cells, frames, or segments (collectively referred to herein as Protocol Data Units (PDUs)) over communication links extending between the network elements. A particular PDU may be handled by multiple network elements and cross multiple communication links as it travels between its source and its destination over the network.
FIG. 1 illustrates an example network architecture in which subscribers 10 are configured to communicate with a central office 12 over links 14. The central office aggregates signals from multiple subscribers and passes the signals onward toward the network 16. There are several technologies that may be used to connect subscribers with the central office, the most prevalent of which is the use of copper twisted pair wires, commonly referred to as subscriber loops. A given subscriber loop may have multiple segments. Other metallic links exist as well, such as metallic links formed using multiple twisted pair wires, and metallic links formed using coaxial cables used to provide cable television signals to cable television customers. Copper links in a communications network, such as subscriber loops and television cables, will be referred to herein as “metallic links.” This defined term is to be understood to encompass links made of copper as well as other metallic materials should such materials be used to form communications networks. The term “metallic link” does not include materials such as optical fibers that are not electrically conductive.
There are many physical layer protocols which may be used to define how data is to be transmitted over a particular metallic link. For example, signals may be formatted according to one of the T1/E1/DS1 protocols, the T3/E3/DS3 protocols, or another of the T/E/DS protocols. Additionally, signals may be formatted using one of Digital Subscriber Line (DSL) protocols, such as Asymmetrical DSL (ADSL), High-bit-rate DSL (HDSL), Rate Adaptive DSL (RDSL), Symmetric DSL (SDSL), or Very high speed DSL (VDSL), although this list is not definitive as other DSL protocols may exist and may be developed in the future. These, and other DSL-based technologies, will be referred to herein collectively as xDSL. Other technologies, such as one of the Data Over Cable Service Interface Specification (DOCSIS) protocols, also adopted as a protocol known as ITU J.112, one of the EuroDOCSIS protocols, or the 802.14 protocol, may also be used to format signals for transmission over cable television metallic links. A given physical layer protocol (layer 1 protocol) instantiated on a given metallic link will be defined herein as a particular “physical medium.”
A subscriber may have a single computer connected to the metallic link 14 or may have a group of computers connected, e.g., over a Local Area Network 18. A network element configured to interface between the subscriber's computer or LAN 18 the metallic link 14 will be referred to herein as Subscriber Premises Equipment (SPE). The computer or local area network may desire to communicate over the metallic link using one or more higher layer (layer 2 or layer 3) protocols, such as Asynchronous Transfer Mode (ATM), Frame Relay, Ethernet, Internet Protocol/Point-to-Point Protocol (IP/PPP), Voice, and Infiniband, depending on the configuration of the local area network and the purpose for the communication. Other layer 2 (link layer) and layer 3 (network layer) protocols exist and this list of protocols is not definitive. Layer 2 and layer 3 protocols will be collectively referred to herein as “services.” Traffic from the subscriber LAN that is required to exit the LAN, as well as voice traffic, will traverse the SPE and be forwarded by the SPE to the central office over the metallic link.
The large number of physical mediums to be supported as well as the large number of services that can be provisioned over the physical mediums may cause problems for the central office. Specifically, the central office needs to provision specific ports on access devices to handle each of the physical mediums. Additionally, each of the ports must be provisioned to handle traffic formatted according to the particular service for that subscriber.
FIG. 2 illustrates one conventional network access device 20 that may be utilized at a central office to interface with various physical mediums. As shown in FIG. 2, a network access device will typically include one or more line interface units 22, each of which acts as a physical interface to a particular type of physical medium. For example, in the illustrated embodiment, there is a set of line interface units configured to interface with metallic links operating using one of the xDSL protocols, a set of line interface units configured to interface with metallic links operating using T1/E1 protocols, a set of line interface units configured to interface with metallic links operating using T3/E3 protocols, and a set of line interface units configured to interface with metallic links operating according to one of the cable television standards, such as DOCSIS. Each line interface unit is associated with a particular port which interconnects the network access device with the metallic link to the subscriber. A given network access device may be configured to interface with any number of physical mediums and may have hundreds or more ports.
Signals received through a line interface unit are conditioned and passed to a framer 24 which removes physical layer specific information from the signals and reassembles frames passed over the subscriber loop. Because the different physical mediums format data differently for transmission, a different framer is required for each type of physical medium supported by the network access device. Reassembled frames are passed to a service mapper 26 configured to reassemble protocol data units to restore the original protocol data units, e.g. packets, delivered by the LAN to the SPE, so that the protocol data units may be handled by the network element and forwarded onto the remainder of the network. Where the network access device is connected to an optical network, the protocol data units are then passed to an optical interface 28 for transmission from the network access device onto the higher communication bandwidth resources afforded by the network 16.
In a conventional network access device, such as the network access device illustrated in FIG. 2, one framer is required for each physical medium supported by the network element. Additionally, for each framer, a set of service specific mappers is required to enable the network access device to map protocol data units onto the particular physical medium. Thus, if a network access device was to support N types of physical layer protocols (e.g. xDSL, T1/E1, T3/E3, and Cable) and M types of services (e.g. ATM, Frame Relay, Ethernet, IP/PPP, Voice, and Infiniband) the network access device would need to implement NxM service mappers. In the embodiment illustrated in FIG. 2, for example, the network access device would need to implement 24 different service mappers. Accordingly, the number of service mappers becomes increasingly large as the number of types of physical mediums supported by a network access device increases and as the number of services supported by a given network element increases.
Additionally, provisioning ports on the network access device becomes increasingly more complex as the number of supported services increases. Specifically, in addition to requiring the network access device to keep track of which ports are to use which physical layer protocol, the network access device is also required to pay attention to the type of service provisioned over that port and to cause the appropriate service mapper to be associated with the port assigned to the particular subscriber. Where the subscriber changes its internal network configuration, for example to migrate from Frame Relay to Ethernet, the network provider is required to reconfigure the port assigned to the customer (possibly manually) to cause traffic from that port to be sent to the new service mapper.
Moreover, certain services cannot be transmitted over certain physical layer protocols. For example, the T1 physical layer protocol requires Point-to-Point Protocol (PPP) to be used as a link-layer protocol. Where the client desires to run Ethernet over a T1 subscriber loop, the SPE is required to terminate the Ethernet connection and reformat the data to be transmitted over the T1 subscriber loop using PPP. The reverse process occurs at the central office whereby the PPP frames are discarded and Ethernet frames are recreated. This creates additional overhead and prevents a true Ethernet network from being formed where the traffic must traverse a T1-provisioned link.
Accordingly, it would be advantageous to alter the characteristics of how communications between subscribers and the central office takes place to reduce the complexity of the network access device and simplify operation to reduce the amount of overhead associated with operating the network access device.